一种多通道注意力机制的lncRNA-miRNA关联预测方法
By incorporating feature representations from different data sources through a multi-channel attention mechanism and contrastive learning method, the problem of non-robust feature representations in existing technologies is solved, and more efficient lncRNA-miRNA association prediction is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEAST FORESTRY UNIV
- Filing Date
- 2024-08-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing methods fail to properly integrate feature representations from multiple data sources and fail to learn the similarities and differences between samples through contrastive learning methods, which may result in the failure to learn more effective and robust feature representations.
A multi-channel attention mechanism is employed to integrate the sequence, expression profiles, and association information of lncRNA and miRNA. The similarity matrix is processed using GCN and Transformer, and contrastive learning methods are combined to capture information interactions, resulting in the final feature representation.
It improves the transparency and interpretability of feature representations, enhances the model's discriminative and generalization abilities, and improves prediction performance.
Smart Images
Figure CN119207579B_ABST